0x79248c897924…79248c86
Ethereum Glamsterdam Tests Expose Payment Failure Flaw on Sepolia
Glamsterdam upgrade tests on Sepolia found transactions reporting success while payments fail, letting NFTs transfer unpaid and withdrawals skip a 2.5% fee.

Outputs
Tests of Ethereum's Glamsterdam upgrade on Sepolia exposed transactions reporting success while internal payments fail.
One scenario delivered NFTs to users without associated payments reaching recipients.
Another scenario let withdrawals bypass a 2.5% cash-out fee.
The flaw stems from proposed changes to gas consumption and storage operations.
No live exploit or stolen funds — the findings are limited to testnet simulations.
Developer simulations of Ethereum's proposed Glamsterdam upgrade on the Sepolia test network have exposed a flaw that allows transactions to report success even when internal payments fail. The findings come from testing conducted ahead of the upgrade's targeted deployment, and they land at a sensitive moment for Ethereum's roadmap, when every protocol-level change gets scrutinized for how it ripples through smart contracts already in production.
What exactly did the tests find?
The simulations surfaced two concrete failure modes. In one, users received NFTs while the associated payments never reached the intended recipients — the transfer of the token asset succeeded, the leg that moved value to the seller did not. In the other, withdrawals bypassed a 2.5% cash-out fee entirely, meaning an application's monetization logic was silently skipped rather than enforced.
Neither scenario involves a live exploit or stolen funds. This is a testnet finding, not an incident on mainnet. But that distinction is exactly why developers treat it seriously: it's the kind of bug you want to catch on Sepolia, not discover after a hard fork ships.
Why does a successful transaction matter so much?
The root cause sits in the proposed changes to gas consumption and storage operations. Under certain conditions, an internal payment can run out of gas while the broader transaction continues to completion and reports success. That inverts a assumption that much of Ethereum's application layer depends on.
Payment applications and smart contracts routinely rely on transaction success as the guarantee that fee collection and fund transfers actually happened. If a transaction can return success while a sub-operation — an internal transfer, a fee deduction — silently fails, that guarantee breaks. Marketplaces that release NFTs on successful payment, platforms that take a percentage cut on withdrawals, escrow-style contracts that settle multiple legs atomically: all of these designs assume the transaction's success status means every step within it succeeded.
For a marketplace, the NFT-without-payment scenario means inventory leaves the seller's control with no compensation arriving. For any protocol charging a 2.5% fee on cash-outs, the second scenario means revenue leakage that the contract cannot detect, because nothing reverts and nothing throws an error the application can catch.
What happens next?
The findings now sit with the developers evaluating Glamsterdam's proposed changes, and the burden is on them to resolve the gas and storage behavior before the upgrade moves toward mainnet activation. The issue is a design-level problem in how the proposed changes interact with sub-calls, not a patch-and-ship bug — which typically means specification work, more testnet rounds, and coordination with application developers whose contracts depend on the current semantics.
The practical takeaway for teams building on Ethereum: audit any contract logic that treats transaction success as proof of payment. The upgrade timeline now depends on how quickly the client teams reconcile the new gas model with the atomicity guarantees the application ecosystem was built on.
via bitcoinke.io (Source)
More from Daniel Okafor
Show full bio
Correspondent covering people, launches and day-to-day practice in crypto and web3 industry at Mempool Brief.
1,514 articles